Research On Collaborative Optimization Of High-speed Railway Train Stop Pattern And Train Schedule Based On Passenger Flow Assignment

With the continuous improvement of high-speed railway network and the gradual improvement of train service level,high-speed railway as a means of transportation is more and more popular.But throughout the transportation industry,the rapid development of civil aviation and the remote leadership of highway industry still make railway enterprises dare not to be underestimated.How to improve the competitiveness of passenger product to meet the needs of passengers to a greater extent is something railway enterprises must consider.In depth discussion of passenger travel choice behavior will provide us with some directions.Therefore,it is very important to describe passenger choice behavior in detail when optimizing the train stop pattern and train schedule directly related to passengers.Based on this background,this paper studies the following contents:(1)Firstly,the passenger choice behavior of high-speed railway is analyzed,and a passenger flow assignment method based on train schedule is proposed.After stating the influencing factors of passenger choice behavior,this paper analyzes and explains that the differences of train arrival and departure time,stop structure and ticket price have an impact on passenger choice behavior.Therefore,it is considered that mutual feedback with passenger flow assignment must be considered in the process of optimization of the train schedule.In the process of passenger flow assignment,considering the behavior of railway passengers purchasing tickets in advance,the passenger flow assignment is further divided into two levels: passenger flow assignment strategy and loading method.The differences of passenger flow assignment results under different passenger flow assignment strategies(user equilibrium and logit)and loading methods(including loading sequence and loading granularity)are compared.(2)Based on the above,a bi-level programming model is established and an algorithm is designed to solve it.The upper model solves the collaborative optimization problem of stop pattern and train schedule,which is solved by improved particle swarm optimization algorithm;the lower model is the passenger flow assignment model.Considering that there is no congestion problem of high-speed trains,this paper introduces the differential pricing instead of congestion as the factor of passenger demand transfer in the lower model,and decides the fare of each route at the same time of passenger flow assignment.In order to compare the effects of different passenger flow assignment strategies,this paper designs two lower level models: passenger flow assignment model based on user equilibrium and improved passenger flow assignment model based on Logit,which are solved by Frank-Wolfe and MSA algorithm respectively.(3)A small case is designed to illustrate the validity of the model,and the results of passenger flow assignment under various passenger flow assignment scenarios are compared and analyzed,as well as the optimization effect of the train schedule.It is found that the scenario with logit strategy can always let more passengers travel on the train with lower travel cost,and when the passenger flow is loaded in batches using the order of ticket purchase,the passenger flow assignment are closer to the real departure of passengers,and at the same time,the train schedule has also been optimized to a certain extent.(4)The upward direction of Beijing-Shanghai high-speed railway is selected as a real case for verification.According to the operation results of the small case,this paper uses logit strategy for passenger flow assignment,and loads the OD passengers in batches according to the ticket purchase order.The optimized train schedule can not only reduce the total travel time of passengers while meeting the needs of passengers,but also effectively reduce the number of train stops and shorten the average travel time of the train,indicating that the optimization performance of the model is good.